In spite of the magnitude of the problem which prostate cancer presents, very little is known regarding the molecular mechanisms underlying prostatic carcinogenesis. It is clear from the recent progress made in colorectal, renal and breast cancer that analysis of familial forms of common human neoplasms can yield tremendous insight into the specific genetic mechanisms in both hereditary and sporadic forms of such cancers. The study of hereditary prostate cancer is a useful index of individual risk and demonstrating that some familial prostate cancer provides the basis for the linkage studies proposed here. The specific aims of this study are to: 1. Identify and recruit multiplex families, from which DNA samples will be obtained for linkage analysis, This effort will build upon our existing family database of over 1300 families, of which 415 are considered informative for this analysis. We have obtained blood from 60 of these pedigrees with a total of 407 individuals and 202 affected. We will collect a minimum of 140 additional pedigrees. Blood has also been collected from affected sibpairs. We will continue to collect with the aim of collecting all 275 identified sibpairs; 2. Undertake a genome wide search for linkage. Having completed an analysis of candidate regions, we will perform a genome wide search for linkage employing approximately 300 simple sequence repeat polymorphisms spaced at 8-10cM intervals throughout the genome. The data will be analyzed for evidence of linkage using the LINKAGE and LODLINK programs. The linkage data will be supplemented with the analysis of affected sibpairs using non-parametric methods; 3. Upon discovery of evidence of linkage, the gene location will be refined using key recombinants, and a fine structure physical map of the identified region will be constructed. 4. Genes within the minimal region will be cloned using a cDNA hybridization selection approach. Candidate genes will be analyzed for mutations in affected individuals as well as tumors. The ultimate goals of this study are to confirm the existence of Mendelian forms of prostate cancer and to identify the responsible gene(s). The identification of genes involved in an inherited form of prostate cancer would be useful for both diagnosis and treatment of groups of men at high risk of disease, and would serve as a model for prostatic carcinogenesis in general.